Virtual Stability Study
Put three batches into ICH chambers and run the clock forward three years. Change the packaging, the degradation pathway and the storage conditions, then get the two answers that matter: has significant change occurred, and what shelf life does ICH Q1E actually support?
- ICH Q1A conditions
- Q1E regression
- 6 fault scenarios
- Printable report
Stability study report
Assay against time
Stability data
This is a teaching simulation. Degradation follows an Arrhenius model with a humidity term and the regression follows the ICH Q1E approach, but the product is invented and no result here applies to a real dossier. Confirm every requirement in ICH Q1A(R2) and Q1E, and set your own specification limits.
Six ways a stability study goes wrong
Run a clean study first and note the shelf life. Then introduce one fault at a time and watch what it costs you — in months of shelf life, and in whether the study is usable at all.
- Chamber excursion. A permanent step in the trend line. The regression never recovers, so a two-week equipment failure can cost a year of claimed shelf life.
- Desiccant omitted. Water content climbs and hydrolysis accelerates. Switch the pathway to hydrolysis-prone and watch accelerated conditions fail outright.
- Light exposure. Assay falls while water content stays normal — that combination tells you it is photolysis, not hydrolysis.
- Humidity probe low. The chamber runs wetter than the record claims, so the data degrades faster than the stated conditions can explain.
- Timepoints pulled late. A quiet steepening of the slope that shortens the shelf life for no scientific reason.
- One batch only. Nothing to pool, a wide confidence interval and a much shorter justifiable claim. The clearest argument for putting three batches on study.
The most valuable experiment here: run the same product in an Alu/Alu blister and in a PVC blister. The packaging decision alone can be worth years of shelf life.
The science behind the chambers
Degradation
Loss of assay follows first-order kinetics with an Arrhenius temperature dependence and a humidity term, which is why a chamber ten degrees warmer degrades a product several times faster:
The humidity the product actually experiences is not the chamber humidity — the packaging stands between them. An Alu/Alu blister lets through almost nothing, a PVC blister a great deal, which is why the packaging decision is a stability decision.
Significant change
ICH Q1A treats a 5% change in assay from the initial value at accelerated conditions as a significant change, along with any failure of specification. When it happens, the intermediate condition must be evaluated and accelerated data can no longer support extrapolation. The simulator applies that test at the six-month accelerated point.
Shelf life by ICH Q1E
Assay is regressed against time. Where the batches behave alike they are pooled; where their slopes differ they cannot be, and the worst batch governs. The shelf life is where the one-sided 95% lower confidence bound on that line reaches the lower limit — not where the line itself does, because the limit must hold with confidence, not on average.
Q1E then caps what you may claim: no more than twice the length of the long-term data, and no more than 12 months beyond it. Twelve months of data supports at most 24 months. That ceiling, not the chemistry, is what usually limits a first submission — move the long-term data slider and watch the proposed shelf life jump as the cap lifts.
Use Copy stability data and paste into the Shelf Life calculator to work the regression through properly.
Designing a stability programme?
We plan stability studies, write the protocols and defend the shelf life in the dossier.